Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    percentage values and comparisons

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    开发者
    0.54
    ร์
    0.53
    Grab
    0.51
     transit
    0.50
    情
    0.49
    Dis
    0.49
    Ling
    0.48
    Vista
    0.48
    <unused60>
    0.48
    Transit
    0.48
    POSITIVE LOGITS
     Fourth
    0.61
     Almost
    0.61
    angering
    0.61
    cdots
    0.59
     compare
    0.59
     presque
    0.58
     Comparable
    0.58
     거의
    0.57
     comparaison
    0.57
     comparer
    0.57
    Act Density 0.016%

    No Known Activations